NPSHA 1

[RS82]

Hopefully somebody could correct me here as I must be doing something wrong. The suction pressure on this vertical pump is 106psig and the vapour pressure is 100psig. The suction pressure was taken off of the centrifugal pump datasheet. Subtracting the vapour pressure from the suction pressure, which I assume takes into account the loss from the pipes and fittings and is the actual pressure going into the pump, 106-100=6psid which is equivalent to 18ft of head, given the S.G. of the fluid is 0.77.

Is it fair to say the NPSHA is 18ft? Reason why I ask is, when the end user was queried on the NPSHA, they said it was only 3ft but confirmed the pressure values they provided were correct???

How so? Also, question#2..is it possible to be have the vapour pressure very close to the suction pressure but still not cavitate as the NPSHA>NPSHR?

 

[TD2K]

The pressure at the suction of the pump above the vapor pressure is 6 psi as you said (I would add the caveat that you should verify the suction pressure in the field rather than assuming the pump data sheet value is correct).

NPSHR curves are in ft of water so you have 6 * 2.31 or nearly 14' of water for NPSHA. Do you know what the NPSHR for the pump is?

 

[BigInch]

Don't confuse cavitation with vaporization of product and trying to pump a gas.

TD2K
NPSHR curves are in feet.
NPSHR curves require head of product being pumped.
Not some equuivalent head of water, unless the product is water.

RS82[/]
Please NOTE! vapor pressures are customarily given as ABSOLUTE pressures.

Pressures must be converted into feet of product.
NPSHA includes atmospheric pressure, or, if the fluid is contained within a closed tank or vessel, it includes the absolute pressure inside that closed tank or vessel

NPSHA = the equivalent head of (Pa - Pv + Pn)
Pa = atomospheric pressure, psiA
Pv = vapor pressure, psiA
Ps = suction pressure, psiG

Density of Water = 62.4 pcf

NPSHA = Ha - Hv + Hn Note! This is at the pump suction inlet.
Ha = atmospheric Head = 14.7 psiA *144/(62.4 * 0.77) = 44.06 ft
Pv = vapor Head = 100 psiA * 144/(62.4 *0.77) = 299.7 ft
Pn = pressure
at pump suction = 106 psiG * 144/(62.4 *0.77) = 317.7 ft

NPSHA = 44.06-299.7+317.7 = 62.04 ft NPSHA

The problem with high vapor pressure products is that when pipe pressure pressure reaches vapor pressure, they vaporize, even though the pump may calculate as haveing sufficient head, if the fluid remained as a liquid. That often happens on the way to the pump, as these products are actually stored at their vapor pressure inside the tank. There's not much head remaining to work with (usually only the head from liquid surface to tank take-off nozzle) before vaporization is occuring.

Also be careful of where the suction pressure is actually taken. If it was really taken at the tank, you must subtract pipe flow losses, but can add any height of fluid level over the pump Centerline, or subtract that distance if fluid level is below pump centerline. If the pump is near the tank, some people think tank pressure is a close enough of an approximation to pump suction pressure.

Since this is a vertical pump, I assume the pressure is suction pressure at the actual intake into the pump. That is apparently 106 psiG or 120.7 psi Absolute. Well above your ABSOLUTE vapor pressure of 100 psiA.

http://www.calculatoredge.com/calc/NPSH.htm

What would you be doing, if you knew that you could not fail?

 

[TD2K]

True Big Inch. NPSHA/NPSHR are in feet of liquid, not feet of water, must have been too late last night.

I'll let the OP comment if the 100 psig vapor pressure is a typo or if he's already converted from psia.

Checking the NPSHA using a suction pressure is a nice way to eliminate the head effect and any line losses, suction strainer, etc. However, any changes in vessel level and/or suction line losses (such as a plugged suction screen) is going to affect the available NPSHA.

 

[BigInch]

2:27 am? Ya...NPSH is best calculated in light of day.

What would you be doing, if you knew that you could not fail?

 

[RS82]

Thanks for the response folks.

The vapour pressure was 114psia which I then converted to gauge.

One more question for you folks..I was told by a pump vendor that even though the suction pressure is very close to the vapour pressure, there is nothing to worry about as there is sufficient NPSHA..if NPSHA is a function of vapour pressure, then how is it sufficient?? Is he wrong?

 

[BigInch]

He's probably thinking that you gave him suction pressure in psig and vapor pressure in psia, thus if equal values, making NPSHA actually 14.7 psia higher than vapor pressure and a much higher head when you divide by the fluid's density... as per my calc above. However he didn't say how close and, if suction pressure gets equal to or lower than bubble point (all in absolute units), that's where close definitely does not count.

What would you be doing, if you knew that you could not fail?

 

[TD2K]

NPSHA is simply the additional pressure you have at the pump suction above the vapor pressure of the fluid. You can see this by taking the traditional NPSHA equation and re-arranging it.

"I was told by a pump vendor that even though the suction pressure is very close to the vapour pressure, there is nothing to worry about as there is sufficient NPSHA"

It's really the margin between NPSHA and NPSHR you need to review. If I have 12' of NPSHA and my pump needs 3' NPSHR, I'm good. If I have 12' of NSPHA and my pump needs 10' NPSHR, I'm going to be worried.

 

[rmw]

I'll worry with you TD2K. I like to have a nice fat margin on NPSH. Most pump vendors lie with respect to their NPSHR curves - or their curves are developed under laboratory conditions and they don't get that in real field conditions.

rmw

 

[GaTechTheron]

RMW,

I would not say that pump manufacturers routinely lie about NPSH3, but rather the topic is widely misunderstood even applications folks among pump manufacturers and data is routinely misrepresented by the end user.

To TD2K's point, the important factor is the difference between NPSH3 and NPSHa because of the way inwhich NPSH3 is actually measured. Because NPSH3 is actually based on a 3% head loss, then technically, cavitation has already begun somewhere between the NPSHa, and NPSH3 values. The inception of cavitation will be just before the differential head of the pump begins to decrease relative to the pump's suction head when the pump company develops their "Knee curves."

RS82, I am not sure if the pump manufacturer did misunderstand the pressure inwhich you provided to him. However, if he meant that the near limiting case of 0.01[ft] - 0.00[ft], means that there you have 0.01[ft] of NPSHa, then the pump manufacturer's understanding of NPSH3 calculation, measurement uncertainty, and design safety margin are clearly flawed. Of course, this is the limiting case.

 

[1gibson]

Keep in mind this is a vertical pump, and everyone is discussing NPSHA relative to the suction nozzle. What is the pump length? You can have sufficient NPSHA for a vertical pump with 0 at the nozzle, provided the pump is long enough.

The only time a pump vendor *might* lie about NPSHR curves is to artificially reduce the NSS so a selection is not rejected due to a specified NSS limit. They *might* say pump needs more NPSHR than it really does, which would make the selection conservative.

If it went the other way, they (pump vendor) would have to deal with the problems through warranty. There is simply no advantage.

 

[rmw]

Early in my career I worked for a company that made seawater desalination equipment. We had 3 saturated fluids to pump, distillate, brine, and often steam condensate. Because the NPSHR determined how high we had to build our equipment above the ground (or pump suction), getting NPSHR right was critical to us.

Therefore we tested all pumps that we intended to use on our product and found out that precious few NPSHR curves were accurate. We learned to routinely add 2-3 feet. Except for Carver Pumps. With the Carver's that we bought if they said 6 ft NPSHR, we built it 6 feet over the pump suction. For the rest we tested and added margin.

Did they lie, or were they just that bad? Carver's results made me think the former but I don't discount the latter.

rmw

 

[BigInch]

Did you use cold water and do your tests according to the standard where HPSHR is determined at a 3% drop in differential head?

What would you be doing, if you knew that you could not fail?

 

[rmw]

I don't remember. I've had a few too many coffee breaks since then.

rmw

PS: I do know that we didn't test for HPSHR at all.

 

[Artisi]

Can't see the point of a manufacturer lying about NPSH - ok they might stretch the truth a little especially if they are aware that the advised NPASHa is well padded etc. As for testing NPSHr under ideal conditions on their test rig - why not - at least you have an accurate starting point for discussion if the "on-site" conditions are wrong (more often than not)

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[BigInch]

They're not stretching the truth. The problem derives from the standard method of determining what value of NPSH is NPSHr. 3% loss is probably 13% too late in many cases not involving cold water as the pumped medium.

What would you be doing, if you knew that you could not fail?

 

[1gibson]

Which is why customers need to specify a margin. Otherwise, the only margin that exists is based on how conservative the manufacturer is with their claims, and of course that is a toss up.